Performance Tuning in ODI 12c

1
votes

I want to migrate more than 10 million of data from source (oracle) to target (oracle) in shorter time in Oracle Data Integrator 12c. I have tried to create multiple scenarios and assign each scenario a million records and run the package of 10 scenarios. Time was reduced but is there any other way so that i can increase the performance of my ODI mapping having more than 10 million records?

I expect a less time for the execution of the mapping for better performance.

1
sqloracleoracle-data-integrator
Please don't measure things using Indian words - James Z
use this link kpipartners.com/blog/bid/149359/… - Roberto Hernandez
We need more details about your mappings. Which knowledge modules (LKM, IKM and potentially CKM) are you using? What is the network architecture between your source and your target? Which steps takes the longest in the operator? Do you have indexes? Do you have partitioning? Are the table statistics correctly gathered? Are you using hints? - JeromeFr
For the loading stage we are using LKM Oracle to Oracle , IKM oracle incremental update, CKM oracle. The steps which takes longer is Update existing rows , we dont have indexes, partitioning, using hints on target. - Arif Ali

1 Answers

-1
votes

Further to add, on the ODI 12c you have option to place the source side indxes, in case the query is performing bad on the source side under the joins option you that option in the physical tab, else you can also deploy the hints in the physical tab. please let me know if these work.